Safety and Preliminary Efficacy of Pembrolizumab Following Transarterial Chemoembolization for Hepatocellular Carcinoma: The PETAL Phase Ib Study

Abstract

Purpose: Transarterial chemoembolization (TACE) may prime adaptive immunity and enhance immunotherapy efficacy. PETAL evaluated safety, preliminary activity of TACE plus pembrolizumab and explored mechanisms of efficacy.

Patients and methods: Patients with liver-confined hepatocellular carcinoma (HCC) were planned to receive up to two rounds of TACE followed by pembrolizumab 200 mg every 21 days commencing 30 days post-TACE until disease progression or unacceptable toxicity for up to 1 year. Primary endpoint was safety, with assessment window of 21 days from pembrolizumab initiation. Secondary endpoints included progression-free survival (PFS) and evaluation of tumor and host determinants of response.

Results: Fifteen patients were included in the safety and efficacy population: 73% had nonviral cirrhosis; median age was 72 years. Child-Pugh class was A in 14 patients. Median tumor size was 4 cm. Ten patients (67%) received pembrolizumab after one TACE; 5 patients after two (33%). Pembrolizumab yielded no synergistic toxicity nor dose-limiting toxicities post-TACE. Treatment-related adverse events occurred in 93% of patients, most commonly skin rash (40%), fatigue, and diarrhea (27%). After a median follow-up of 38.5 months, objective response rate 12 weeks post-TACE was 53%. PFS rate at 12 weeks was 93% and median PFS was 8.95 months [95% confidence interval (CI): 7.30-NE (not estimable)]. Median duration of response was 7.3 months (95% CI: 6.3-8.3). Median overall survival was 33.5 months (95% CI: 11.6-NE). Dynamic changes in peripheral T-cell subsets, circulating tumor DNA, serum metabolites, and in stool bacterial profiles highlight potential mechanisms of action of multimodal therapy.

Conclusions: TACE plus pembrolizumab was tolerable with no evidence of synergistic toxicity, encouraging further clinical development of immunotherapy alongside TACE.

Figure 1.

A, Study flow chart. B, Swimmer's plot of study participants receiving TACE plus pembrolizumab. Each bar represents one subject in the study. Depth, duration of response, dates of radiologic response or progressive disease, and presence of ongoing response are indicated accordingly. C, Kaplan–Meier curve illustrating the PFS of the study population. D, Kaplan–Meier curve illustrating the OS of the study population.

Figure 2.

A, Dynamic changes of ctDNA concentration at various study timepoints and relationship with response to treatment. B, Distribution of individual mutations across pretreatment samples (n = 28). C, Oncoplot illustrating the distribution and type of mutations identified in pretreatment plasma samples of patients recruited to the study. Each column represents a sample and each row a different gene. The left barplot illustrates the VAF pertaining to each sample, while the right barplot has the frequency of mutations in each gene. The top bar plot estimates the tumor mutational burden (TMB) calculated on the basis of number of mutations per megabase of sequenced genome. Samples are ordered by the most mutated genes. D, Distribution of the top 10 T-cell receptor rearrangements as measured by productive frequency prior to pembrolizumab start in responders (n = 6) and nonresponders (n = 7; n = 2 not available). E, Productive Simpson clonality was significantly higher in responders before commencement of systemic therapy. *, P < 0.05; **, P < 0.01.

Figure 3.

A, Heat map illustrating relative representation of individual cell types as assayed using CyToF across the various study timepoints (screening, cycle 1 week 0, that is, post-TACE, and EOT) stratified on the basis of the achievement of response 12 weeks following TACE plus pembrolizumab combination. Hierarchically clustered heat map displays mean frequencies of lymphocyte subsets per timepoint and response to therapy, visualized by z-score–based coloring after column normalization. B, Tc2 cell frequency differed between screening and EOT (n = 8, paired samples). C, Tc1 cell frequency before ICI therapy was different between responders (n = 6) and nonresponders (n = 7). D, B-cell frequency differed at end of therapy between responders (n = 6) and nonresponders (n = 7). *, P < 0.05.

Figure 4.

A, tSNE visualization of CD45⁺CD3⁺CD19⁻ T lymphocytes pooled (top), divided by response to therapy (middle), and in addition by timepoint (bottom). Each dot corresponds to a single cell. Representative data of 4 patients, subsampled up to 15,000 per sample; analysis of 157,543 cells in total. B, Overlay plot indicating the localization of Tc1 and Tc2 cells on the tSNE map. C, T-cell marker expression is visualized on the tSNE map by color heat map. D, T cells were clustered on the basis of their high-dimensional expression profile using FlowSOM. Clusters are visualized by the indicated colors on the tSNE map. E, Cluster abundance of c07 and c08 in responder and nonresponder patients. Data are represented as mean with SD and analyzed by paired t test. F, Hierarchically clustered heat map indicating mean marker expression of T-cell differentiation, activation, and exhaustion markers by each cluster and including manually gated Tc1 and Tc2 subsets as in Fig. 3. Expression is visualized by z-score as indicated; data were scaled per column. *, P < 0.05; **, P < 0.01.